From the Department of Public Health Sciences Karolinska Institutet, Stockholm, Sweden
OBESITY AND STIGMA
STUDIES ON CHILDREN, ADULTS AND HEALTH CARE PROFESSIONALS
Lena M Hansson
Stockholm 2010
All previously published papers were reproduced with kind permission from the publishers.
Published by Karolinska Institutet. Printed by US-AB, Stockholm
© Lena M Hansson, 2010 ISBN 978-91-7409-847-1
We pretend we have open minds, yet we judge so fast How can I change a million minds in this lifetime I don't fit in their descriptions - though I try What is it I have to do
Would I be Good Enough For You?
Jay Sean
To my mother
ABSTRACT
Obesity may not only pose a threat to an individual’s physical health but may also have socioeconomic and psychosocial consequences. Stigma is assumed to be a common experience among individuals with obesity and is also suggested to be one of the major contributors to health disparities. The overall aim of this thesis was to study obesity and stigma from children’s, adults’ and health care professionals’ perspectives in Sweden.
A quantitative design with statistical analyses was used in Study I, II and III. These studies included national random samples of approximately 1400 10-year-old children and their parents and about 2600 men and women between 25 and 64 years of age. A qualitative descriptive design involving a phenomenographic analysis was used in Study IV. This study included a strategic sample of 10 general practitioners (GPs) and 10 district nurses (DNs) from primary health care centres in Stockholm County.
Children were more likely to be prejudiced against obesity than against average weight and thinness. The likelihood of children being prejudiced was found to vary with children’s sex and with the rated figure’s sex and body size. Children with high socioeconomic status (SES) were more likely to be prejudiced against a target with obesity than children with low SES (Study I).
Boys’ lower body esteem predicted a higher level of stereotyping of a girl figure with obesity, whereas girls’ body esteem could not explain the variation in girl obesity stereotypes. Parents’
strong beliefs about the controllability of weight and larger body size both independently predicted a higher level of obesity stereotypes in their children (Study II). Women with severe obesity reported discrimination more often than normal weight women and this was documented in all investigated contexts, i.e. workplace, health care and interpersonal encounters. Reports of discrimination among men seemed to depend on both the context and the level of obesity. Insulting treatment by physicians and nurses, and also experiencing inferior medical care than others and avoiding care due to fear of being mistreated, were most common among individuals with severe obesity (Study III). Even though health care professionals themselves did not explicitly expose negative attitudes towards obesity their conceptions of obesity were centred almost exclusively on lifestyle behaviours as causes and remedies. Successful encounters with individuals with obesity in primary care were conceived to rely a great deal on patient attitudes (motivation to change lifestyle, evasive behaviour, trusting in care, lack of self-confidence). However, the importance of both organizational and staff aspects for improving provider-patient encounters regarding obesity was stressed by both GPs and DNs. The conception that primary care is not an entirely appropriate setting for approaching obesity applied to both professional groups, but especially GPs (Study IV).
In conclusion, the findings in this thesis support the idea that obesity is a stigmatizing attribute in Swedish society; however, different social contexts and social identities condition the stigmatization of people with obesity. The results presented in this thesis can, for instance, contribute to new ideas about how to reduce stereotypes in children. The findings and knowledge gained from health care professionals’ conceptions of their encounters with patients with obesity may also be useful in the continuing efforts to improve care of these patients.
Keywords: obesity; obesity (attitudes towards); stigma; stereotyping; prejudice; bias; children; adults;
gender; primary care; health professionals; weight management; qualitative research
SVENSK SAMMANFATTNING
Fetma innebär inte bara fysiska och medicinska hälsorisker utan kan också resultera i socioekonomiska och psykosociala konsekvenser. Stigmatisering av personer med fetma anses relativt vanligt och stigmatisering i sin tur anses vara en bidragande orsak till hälsoskillnader i en befolkning. Det övergripande syftet med denna avhandling var att undersöka fetma och stigmatisering utifrån barns, vuxnas och hälsopersonals perspektiv.
En deskriptiv kvantitativ design med statistiska analyser användes i Studie I, II och III. Dessa studier inkluderade cirka 1400 10-åriga barn samt en förälder och cirka 2600 män och kvinnor i åldrarna 25-64 år från hela Sverige. En deskriptiv kvalitativ design med en fenomenografisk analys användes i Studie IV. Denna studie bestod av 10 allmänläkare och 10 distriktssköterskor från primärvården i Stockholms län.
Det var signifikant större sannolikhet att barn hade fördomar om fetma än om normalvikt eller undervikt. Variationen i barns fördomar kunde förklaras med barnets kön och kroppsstorlek och kön på den figur som bedömdes. Barn från en hög socioekonomisk bakgrund rapporterade i högre utsträckning fördomar om fetma än barn från låg socioekonomisk bakgrund (Studie I).
Ju mer negativ kroppsuppfattning hos pojkarna desto fler stereotyper hade de om en flickfigur med fetma, medan flickors kroppsuppfattning och stereotyper om en flickfigur med fetma inte var associerade. Om föräldrar hade en stark tro på personligt ansvar för fetma så angav barnen fler stereotyper om fetma, medan större kroppsstorlek hos föräldrar betydde att barnet angav färre stereotyper (Studie II). Kvinnor med svår fetma rapporterade oftare diskriminering än kvinnor med normalvikt. Detta samband fanns i alla undersökta sammanhang, det vill säga, arbetslivet, hälso- och sjukvården och i interpersonella möten. Diskriminering bland män var mer beroende av sammanhang och graden av fetma. Negativa upplevelser från möten med läkare och sjuksköterskor samt upplevelse att på orättvisa grunder blivit vägrad eller fått sämre vård än andra samt att ha undvikit vård på grund av rädsla för att bli utsatt för kränkande behandling var vanligast bland personer med svår fetma (Studie III). Även om hälsopersonal inte uppenbart uttryckte negativa attityder till fetma så var deras uppfattningar i stor utsträckning inriktade på livsstilsfaktorer som orsaker till fetma och lösningen på densamma.
Hälsopersonal hade uppfattningen att patientens attityd (motivation till förändring, undvikande beteende, förlita sig på medicinsk vård, saknar självförtroende) var en viktig aspekt för ett lyckosamt möte i en primärvårdskontext, men att även såväl organisatoriska som personella faktorer var betydelsefulla för en väl fungerande vård för dessa patienter.
Uppfattningen att primärvården inte är en alldeles självklar arena för att ta hand om fetma fanns också hos båda professionerna, men speciellt hos allmänläkarna (Studie IV).
Sammanfattningsvis så visar resultaten från denna avhandling att fetma är ett stigmatiserande attribut i det svenska samhället. Däremot verkar den sociala kontexten och den sociala identiteten ha betydelse för stigmatisering av personer med fetma. Vidare kan resultaten från denna avhandling bidra med nya idéer om hur stereotyper om fetma bland barn kan minskas.
Kunskapen om hälsopersonals skilda uppfattningar om mötet med patienter med fetma kan förhoppningsvis också användas för att förbättra vården för dessa patienter.
LIST OF PUBLICATIONS
This thesis is based on the following papers, which will be referred to by their Roman numbers.
I. Hansson LM, Karnehed N, Tynelius P, Rasmussen F. Prejudice against obesity among 10-year-olds: a nationwide population-based study. Acta Paediatrica 2009;98(7):1176-1182.
II. Hansson LM, Rasmussen F. Predictors of 10-year-olds’obesity stereotypes: A population-based study. International Journal of Pediatric Obesity 2010;5(1):25-33.
III. Hansson LM, Näslund E, Rasmussen F. Perceived discrimination among men and women with normal weight and obesity. A population-based study from Sweden.
In print, Scandinavian Journal of Public Health, SAGE publications.
IV. Hansson LM, Rasmussen F, Ahlström G. General practitioners’ and district nurses’
conceptions of the encounter with patients with obesity in primary health care.
Manuscript submitted.
ABBREVIATIONS
BE Body esteem
BESAA Body esteem scale for adolescents and adults
BMI Body mass index
CI Confidence interval
DNs District nurses
GPs General practitioners
LOUISE Longitudinal database of education, income and occupation
OR Odds ratio
PHC Population and housing census
ULF The Swedish survey of living conditions RTP Register of the total population
SES Socioeconomic status
WHO World Health Organization
TABLE OF CONTENTS
1 INTRODUCTION... 1
2 BACKGROUND ... 2
2.1 The stigma concept... 2
2.2 Children’s stereotype and prejudice formation ... 4
2.3 Theory of obesity stigma ... 6
2.4 Measuring obesity stigma ... 8
2.5 Prevalence and aetiology of obesity ... 9
2.6 Measuring obesity ... 10
2.7 Obesity stigma – Children... 11
2.8 Obesity stigma – Adults ... 15
2.9 Consequences of obesity stigma ... 18
3 AIM OF THE THESIS ... 19
4 METHODS... 20
4.1 Study design ... 20
4.2 Participants ... 21
4.3 Procedure ... 21
4.4 Data collection ... 22
4.5 Data analyses... 26
4.6 Ethical considerations ... 28
5 RESULTS... 30
5.1 Obesity – a stigmatizing attribute... 30
5.2 Moderators of obesity stigma ... 31
5.3 Obesity and encounters in health care... 34
6 DISCUSSION... 37
6.1 Main findings ... 37
6.2 Findings in relation to previous research and theory ... 37
6.3 What mechanism or theory is most probable... 50
6.4 Methodological considerations ... 50
6.5 Concluding remarks and future directions... 58
7 ACKNOWLEDGEMENTS ... 60
8 REFERENCES... 62
1 INTRODUCTION
Research on stigmatization of people with obesity has been quite extensively performed in the United States (US). Studies have included children and adolescents as well as adults, and have covered different settings. Reviews suggest that stigmatization is pervasive in work-life, health care and education, and also in the media and close interpersonal relationships (Puhl & Heuer, 2009; Puhl & Latner, 2007). There is, however, more work to be done in this area. For instance, population-based studies have seldom been performed, but instead research has mostly been based on clinical or convenience samples. Questions also remain as to whether gender or socioeconomic position are important moderators of obesity stigmatization, and – most importantly – how children form stereotypes about obesity is not entirely known. Even though the evidence is strong for the presence of negative attitudes towards obesity among health care professionals; limited evidence is provided for actual mistreatment of people with obesity in health care. Furthermore, although not investigated in the present thesis, the impact of obesity stigmatization on psychological and physical health has received little attention. Available research suggests that obesity stigma increases the risk of depression, low self-esteem, poor body image, maladaptive eating behaviours and exercise avoidance, and possibly also further weight gain. Research has also found low effectiveness in improving attitudes towards obesity in both children and adults (Danielsdottir, O'Brien & Ciao, 2010). Because stigma is an important contributor to health disparities (Link & Phelan, 2006), better knowledge about the origin, development and spread of obesity stigmatization is needed.
Only very recently has research on obesity stigmatization in countries and cultures other than the US increased, and to the best of my knowledge the published research from Sweden remains scanty. There are, however, studies investigating associations between body mass index (BMI) status and social consequences, which show that young men with obesity have lower educational attainment (Karnehed, Rasmussen, Hemmingsson & Tynelius, 2006), and more often show downward then upward social mobility (Karnehed, Rasmussen, Hemmingsson & Tynelius, 2008), than their normal weight counterparts. Also, adolescent females with overweight have been shown to occupy lower social positions than normal weight females 14 years later (Hammarström & Janlert, 2005). These studies can only speculate about stigmatization as the cause of disadvantaged outcomes. A review of 1925 articles in daily newspapers between 1997 and 2001, however, demonstrates stigmatization of obesity in a Swedish media context (Sandberg, 2007). There is a rising prevalence of obesity both internationally and in Sweden (Neovius, Teixeira-Pinto & Rasmussen, 2008). Also, it is claimed that obesity stigma in part has its origin in cultures that value independence and focus on personal responsibility for life outcomes, in combination with the notion that thinness and success go hand-in-hand (Crandall & Martinez, 1996). These are two reasons further to investigate this complex issue. In this thesis population-based samples of children (and their parents) and adults, as well as a strategic sample of health care professionals, were used to study attitudes, stereotypes, prejudice and discrimination in relation to obesity.
2 BACKGROUND
2.1 THE STIGMA CONCEPT
Contemporary concepts of stigma originate from Goffman’s (1963) classic book “Stigma: Notes on the management of spoiled identity”. According to Goffman (1963, p. 3) the definition of stigma is “an attribute that is deeply discrediting which reduces the individual from a whole and usual person to a tainted, discounted one”. Later Crocker et al. (1998, p. 505) proposed that stigmatization occurs when a person possesses (or is believed to possess) “some attribute or characteristic that conveys a social identity that is devalued in a particular social context”.
Stigma can be seen as a relationship between the attribute and a stereotype, and the stereotype becomes the basis for excluding or avoiding individuals with that attribute (Link & Phelan, 2001). Such stereotypes are generally known by a majority of members in a culture, including the stigmatized individuals themselves. By 10 years of age, most children are aware of cultural stereotypes of different groups in society, and children who are stigmatized seem to be aware of these stereotypes at even younger ages (Major & O'Brien, 2005).
The definitions above are in agreement that stigma consists of an attribute that marks people as different and leads to devaluation, and also that stigma is dependent on relationship and context. Link and Phelan (Link & Phelan, 2001) added the component of discrimination to the concept of stigma, which refers to “an unfair or unjustified difference in behaviour that systematically disadvantages members of another group” (Dovidio, Penner, Albrecht, Norton, Gaertner & Shelton, 2008, p. 479). They also included the component of power. Even though both groups; those that are in possession of power and those that are not, may stereotype and discriminate against another, the former group has access to resources and the power to set norms in society. The dependence of stigma on power, when it comes to obesity, may be difficult to apprehend; however, it is evident that there is a power difference between people who have obesity and those who do not.
Mainly, stigma is perceived as a social process with multiple dimensions. In brief, stigma exists when the interrelated components of categorization, stereotyping, separation, status loss and discrimination co-occur in a power situation (Link & Phelan, 2001). In general, stigma places more emphasis on the target, whereas the very closely related concept of prejudice focuses more on processes in perpetrators. In the seminal work by Allport (1954, p. 9) a definition of prejudice, in terms of ethnic prejudice, was put forward: “Ethnic prejudice is an antipathy based upon a faulty and inflexible generalization”. Prejudice seems a little bit narrower in scope than stigma, but when the causes and consequences of prejudice are included, the concepts are similar (Phelan, Link & Dovidio, 2008). However, it has been suggested by several researchers that stigma refers to a broader process, which includes many components, and that prejudice refers to more attitudinal components. An additional term that has been extensively used is bias, more specifically weight bias in regard to obesity stigma. Bias is a broad term that encompasses any combination of negative thoughts, feelings and behaviours (Brownell, Puhl, Schwartz & Rudd, 2005). But, so as not to confuse the reader, the term bias has been scarcely used in the present thesis, because of its common usage in epidemiological research when discussing measurement error.
Phelan et al. (Phelan et al., 2008) attribute three functions to stigma, which indicate the sources and reasons for it. These are exploitation and domination (keeping people down), norm enforcement (keeping people in), and disease avoidance (keeping people away).
Exploitation/domination reflects people’s desire to maintain advantage; that is, for dominant groups to have more resources and power, some other groups have to have less. Stigma against low socioeconomic status (SES) groups, ethnic minorities and women may have its origin in this function. Conforming to social norms in society is perceived to be essential, and those who do not are blamed for lacking character or morality. This type of stigma applies to behaviour and identities that are perceived as voluntary, e.g. criminal behaviour, non-normative sexual behaviour, and obesity. Norms and un-acceptable behaviours in society are often set by dominant groups; however, these groups will not profit from the stigma process as seen in its exploitation/domination function. The function of disease avoidance may be difficult to explain as an entirely social construct, and it has therefore been suggested that evolutionary theory is better suited to this type of stigma. However, this function seems only to apply to those illnesses or physical deviations that are visible (Phelan et al., 2008). Human preferences for facial symmetry (Grammar & Thornhill, 1994), which can be seen very early in life and across cultures, may be evidence for the existence of this stigma function. It is suggested that the attributes an individual carries act as signals. Humans react cognitively to these attributes in order to avoid social contact with individuals who would jeopardize their survival (Major &
O'Brien, 2005). Documented emotional reactions (fear, disgust, nausea) are, it is suggested, to be regarded as reflecting this stigma’s past function (as a social signal of pathogen infection) rather than its function in current contemporary society. So, even if humans have a reactive
“behavioural immune system”, cultural values and beliefs have a role as to play with regard to which types of stereotypes are attached to the stigmatized individual (Major & O'Brien, 2005).
Obvious and overt prejudice has declined as society has changed and been replaced by “modern prejudice”, which is more subtle and hidden, and which often includes contradictory feelings and opinions (Dovidio et al., 2008). Attitudes and stigma/prejudice are strongly related, but sometimes individuals act in a discriminatory manner against a group while not necessarily showing evidence of negative attitudes. An attitude is thought to have an affective and a cognitive component (Crocker et al., 1998). The affective component includes, for instance, feelings of distaste, dislike, disgust or even fear, while the cognitive component is what we refer to as stereotypical, namely “a cognitive structure that contains the perceiver’s knowledge, beliefs, and expectancies about some human group” (Bigler & Liben, 2006, p. 42).
Stereotyping is assumed to involve two processes, one automatic or unconscious, the other controlled. The automatic process is also referred to in terms of implicit attitudes (not investigated in the present thesis), and the controlled process as involving explicit attitudes (Bigler & Liben, 2006). Implicit attitudes are regarded as having evolutionary roots or as being acquired early in childhood through repeated messages. Such attitudes may continue to influence adult behaviours without them being aware of their biased reactions. This is one of the reasons why it is important to understand the origins of children’s stereotypes and prejudice. Explicit attitudes are suggested to be more recently and consciously constructed attitudes (Wilson, Lindsey & Schooler, 2000). Explicit attitudes are thought to predict
deliberative behaviours, while implicit attitudes better predict spontaneous forms of behaviours (Bessenoff & Sherman, 2000). However, there is an ongoing debate in psychology research over the use of attitude and stereotype measures, implicit or explicit, as constructs of real behaviour. To establish links between attitude/stereotype measures and behaviours is a problem in the whole stigma arena (Phelan et al., 2008). One thing that makes such relationships even more difficult to investigate is the fact that our behaviours can also influence our attitudes and stereotypes; thus, it is likely that these different components of stigma are to some extent reciprocally related (Dovidio et al., 2008).
2.2 CHILDREN’S STEREOTYPE AND PREJUDICE FORMATION
Previously, social learning theory has served as a basis for general understanding of children’s stereotyping. The theory emphasizes the cultural and social context of the child and assumes that stereotypes are acquired through observational learning. However, the theory overlooks the developmental process and assumes that the formation of stereotypes is the same across all ages. But, given the evidence that cognitive skills affect the construction of social stereotypes and their meanings (Aboud, 2003), Bigler and Liben (2006) have instead proposed a developmental inter-group theory of social stereotypes and prejudice. Both the environment in which children are raised and the child’s way of interpreting and interacting with the environment is essential for the development of social stereotyping.
Developmental inter-group theory (see Bigler & Liben, 2006) describes four components that are important in the formation and maintenance of social stereotypes:
establishment of the psychological salience of person attributes; categorization of encountered individuals along a salient dimension; development of stereotypes concerning salient social groups; and, application of a stereotype filter when individuals are encountered. The establishment of psychological salience of social groups in children are thought to rest on the awareness of difference (normal weight as opposed to obesity), and is also dependent on the proportional group size (with social categories becoming more salient when group size is unequal). Explicit labelling and use in a society is also important; that is, children develop stereotypes on the basis of those characteristics that the society deems as important for social categorization. This process is, though, presumed to be constructive rather than socially learned. Children construct beliefs about groups based on, for instance, adults’ cues, which may be very subtle. There are also implicit processes, such as segregation, that lead children to view some social categories as important. Children note, for instance, similarities among those who live and work together, and this encourages them to construct beliefs about group differences. Thus, during the socialization process, children develop a set of expectations about people’s physical appearance, manners and behaviours.
Children have an innate drive to classify things in their surroundings, so as to structure knowledge and reduce the complexity of the world; even at the age of three, they are assumed to be able to perform social categorization along a single dimension, and thus show stereotyping and prejudice. For instance, young children show strong same-sex preferences, even though cultural messages about gender traits differ. Later on (by ages 7-8 years), children have obtained multiple classification skills that enable them to sort along two dimensions; thus,
a child who has difficulty in understanding that the same individual can be a member of two groups at the same time will show higher levels of distortion and forgetting when approached with information that counteracts a previously formed stereotype. Cognitive skills that may be relevant to the formation of stereotypes are those associated with perspective-taking (ability to understand another’s thoughts or emotions) and making probability judgments (imagine alternative outcomes of a problem). By the age of 5-8 years, children are posited to peak in their prejudice, since they prefer their in-group to out-groups, and also because they have only just started to realize that people can have perspectives different from their own. This inter-group process is not static, and cognitive processes are likely to affect the formation of stereotypes about social groups. From ages 8-10 and onwards, children will increase in their appreciation of others’ perspectives, and this should lead to lower prejudice. By the age of 10-15, more sophisticated perspective-taking skills are developed, and it is suggested that these are linked to moral reasoning (Kohlberg, 2008). Children’s ethical and moral reasoning and perspective- taking skills may thus in part determine a child’s response to human differences.
The process of categorization is believed to produce constructive-cognitive developmental processes. This process will help in attributing meanings to social groups in the forms of beliefs and affects. According to Bigler and Liben (2006), four factors are regarded as important: essentialism, in-group bias, explicit attribution and group-attribute co-variation, and implicit attribution. Essentialism is the belief that members of a group share the same qualities (those that are similar on one dimension, e.g. obesity, are similar on other dimensions too, e.g.
lazy). In-group bias is produced when children view the group to which they belong as superior to another group, which is also called egocentrism, reflecting a sort of fixation on one’s in-group perspective. Self-esteem has been shown to be related to in-group bias, especially in elementary-school children, and it is suggested that they generalize their positive or negative feelings of themselves to all other in-group members. Thus, variation in children’s self-esteem is believed to be important for the formation of stereotypes. Explicit attribution has to do with labelling and making propositional statements about social groups. Direct teaching may play only a minor role because adults today are most likely to suppress their prejudices. Some explicit attributions may be socially unacceptable and therefore are not expressed, whereas others remain acceptable; for instance, laziness is a stereotypical feature readily attributed by both children and adults to people with obesity. However, if explicit attributions among adults are rare, it is likely that children make them to a much greater extent, and peers may therefore be an important source of stereotypes. Children may also notice that some human characteristics are correlated with specific attributes, and that adults treat individuals differently according to which social group they are perceived to belong. Adults’ non-verbal behaviours are also things that children try to interpret and give meaning. Such implicit behaviours and group attributions may form children’s stereotypes, and because at least young children’s stereotypes are more likely to work as rigid rules, they would, to a greater degree, generalize specific attributes to all members of a particular social group.
Thus, it is assumed that children’s social stereotypes are created through cognitive processes about social categories rather than being copied directly from those held by individuals in their surroundings. It is proposed that such cognitive processes are also used as
filters when new information is about to be handled, and that this new information is often ignored and distorted so as not to counteract existing beliefs. Some adult attitudes and behavioural responses may therefore have their origin in these cognitive processes.
2.3 THEORY OF OBESITY STIGMA
A great deal of research has considered specific stigmas, and knowledge has increased about cognitive, affective, and behavioural components of stigma (attitudes, stereotypes and discrimination). However, obesity stigma as such has scarcely been considered in existing theories, and there are few overarching models of stigma (Bigler & Liben, 2006; Phelan et al., 2008). Some of the research that does exist has tried to explain obesity stigma from the perspectives of social learning, social identity or attribution of blame, but it has also been suggested that psychological theories may not be sufficient fully to capture the phenomenon (Puhl & Brownell, 2003).
In any case, one of the reasons why obesity stigma seems to be such a negative stigma is because it is visible and perceived to be under personal control (Crandall, 1994; DeJong, 1980;
Weiner, Perry & Magnusson, 1988). The more people think that weight is a function of willpower, exercise and diet, the more negative attitudes will be expressed. Attribution theory suggests that people make judgments about the causes of people’s outcomes. In the case of obesity, the individuals themselves are held responsible for their condition, which is attributed to a lack of willpower or laziness (Crandall, D'Anello, Sakalli, Lazarus, Wieczorkowska &
Feather, 2001). Those who believe that obesity is the result of lack of impulse control and other personal shortcomings (internal causes) can be expected to be more likely to express negative attitudes towards obese individuals than those who attribute obesity to medical, heredity or environmental factors (external causes). Attribution theory would, thus, correspond to the stigma function of norm enforcement.
Attributions are also used as justifications for stigma, and therefore such attributions can work in both directions; that is, attributions can both legitimize and cause stigma (Crandall
& Eshleman, 2003). Stigma and prejudice are often suppressed in today’s society, but a suppressed prejudice may be expressed if it can be justified. It has been suggested that attributions of controllability of weight, and especially controllability of obesity, originate from political, economic and or social ideologies that share the common world view that the more one believes in individual responsibility for life outcomes, the more prejudice against obese individuals will be expressed. Attributions predict prejudice in general, but are more important in individualistic than in collectivistic countries (Crandall et al., 2001). Cross-cultural studies also show that prejudice against obesity only emerges if there is both a cultural preference for thinness and a belief that weight is under personal control (Crandall & Martinez, 1996). Thus, attributions may not entirely explain obesity stereotypes because there are social and cultural differences that are related to body weight.
There is considerable evidence that our own attitudes, beliefs and behaviours are influenced by our perceptions of the attitudes, beliefs and behaviours of individuals important to us. According to group-norm theory, stigma can develop through the socialization process (Sechrist & Stangor, 2005). Stereotypes are often shared by members of the same group, and
stereotypes are important in social interaction. Stereotypes are communicated constantly in everyday interactions between peers, children and parents, and through the media, etc.
Knowing that others relevant to us have similar attitudes function as a form of validation, and the social consensus that is established in interactions give individuals a sense of belonging and confidence that their attitudes are shared by others. Research shows, for instance, that when people are told that others hold more favourable views of the obese, they express more positive attitudes after having been given such information, and even more so when information comes from a valued in-group (Puhl, Schwartz & Brownell, 2005). The social consensus model explains stigma from a social constructivist perspective, where the emphasis is on how one perceives the stigmatizing attitudes of others (Sechrist & Stangor, 2005). Thus, individuals may be more likely to form beliefs about people with obesity from their perceptions of others’ beliefs rather than from their own interactions with obese individuals.
Social and cultural norms about body size are very likely to be an influential source in the formation of obesity stereotypes and stigma. Socio-cultural theory suggests that a number of socio-cultural influences (peers, family, media) transmit societal standards, e.g. in the case of beauty (Smolak, 2004; Thompson, Coovert, Richards, Johnson & Cattarin, 1995). This model emphasizes the desirability of an unrealistic level of thinness in people, especially among adolescent girls and young women. Thinness has long been associated with success and beauty (Grogan, 2008b), and this is regarded primarily as being achieved by diet and exercise. This image of thin people being successful seems to apply mostly to females, while males’ image of success is more variable, even though muscular and slim seem to be the preferred norm. The stronger the belief that thinness is associated with success, the more likely it is that the thin ideal is also used when judging peers or others (Davison & Birch, 2004; Klaczynski, Daniel &
Keller, 2009). Experts in the field of medicine also stress the importance of thinness, or at least average weight to stay healthy, and also the protective role it plays against lifestyle diseases (Muennig, 2008). Historical changes in beauty ideals (Grogan, 2008b) and the focus on the importance of average weight for health, in conjunction with the media’s and the diet industry’s emphasis on dieting and exercise for weight reduction, may therefore have had the unintentional effect of increasing obesity stigma (Latner & Stunkard, 2003).
Social norms in society are most often set by dominant and powerful societal groups and, although there are few studies that confirm that attitudes are more negative towards obesity among higher social groups, there are still findings to support this notion. In one study, children from high SES schools assigned fewer positive adjectives to an obese figure compared with children from low SES schools (Wardle, Volz & Golding, 1995), while, in another, fathers of 9-year-old girls with a higher education and income were found to be more likely to endorse obesity stereotypes (Davison & Birch, 2004). However, highly educated individuals have also reported less negative attitudes than low educated individuals (Hilbert, Rief & Braehler, 2008).
Furthermore, the prevalence of BMI is lower and body dissatisfaction seems to be higher among high SES individuals than low SES individuals, and as women’s level of education has increased over time, the body ideal has become more slender (Grogan, 2008a; McLaren &
Godley, 2008). Socioeconomic differences in attitudes and beliefs about healthy lifestyle have also been documented (Wardle & Steptoe, 2003). This highlights that appearance investment
and attitudes about lifestyle can be group-level attributes, not just individual-level characteristics.
Obesity stereotypes emerge early, at around the age of three (Cramer & Steinwert, 1998), and seem to become stronger with age (Brylinsky & Moore, 1994; Klaczynski, 2008; Klaczynski et al., 2009). This might be interpreted in terms of children being gradually exposed to cultural messages about obesity and the ideal of thinness in society, which would be consistent with social learning theory. However, recent studies have shown that both adults and children react to individuals with obesity as if they were carrying some kind of disease (Klaczynski, 2008;
Park, Schaller & Crandall, 2007); that is, it is suggested that obesity serves as a heuristic cue (a simple rule that helps people respond quickly to complex problems or incomplete information) for pathogen infection. Rejection of obesity is expressed to the same extent in children within cultures with both a low and a high prevalence of obesity, and with media that do and do not display obese people negatively. This suggests that obesity stigma is not only a function of social learning (Klaczynski, 2008). People are often unable to say how they have made judgments on physical appearance, indicating that implicit attitudes are important.
Furthermore, children seem not always able to report the reasons for their reactions, but intuitively they would feel that they are correct (Major & O'Brien, 2005). This sense that there is something wrong and unfamiliar about individuals with obesity could explain the avoidance of people with obesity (Klaczynski, 2008). Even though the disease avoidance mechanism may have a role in the stigmatizing of obesity, developmental and cultural aspects seem important as well. This mechanism may later on be transformed into a set of beliefs about obese people, i.e. obesity stereotypes, which strengthen with age. During the child’s developmental stages, these stereotypes are also likely to be reinforced by, for instance, the media, which present obese characters in a biased way (Greenberg, Eastin, Hofschire, Lachlan & Brownell, 2003), parents who relay the idea that obese individuals are unsuccessful (Adams, Hicken & Salehi, 1988), and a society that idealizes a thin appearance (Thompson & Heinberg, 1999).
Consequently, children will be exposed to a variety of cues to which they will respond by trying to construct meaning.
2.4 MEASURING OBESITY STIGMA
The assessment of attitudes, stereotypes, and discrimination in general has a long history, but measurements of these components in relation to obesity are rather new. However, research has used diverse techniques to capture stigma against people with obesity, which have included survey methods, questionnaires with experimental manipulations, laboratory experiments and field studies. Explicit attitudes (self-reported and endorsed) and implicit attitudes (outside conscious control and awareness), and also behavioural reactions and personal experiences of stigma, have been investigated.
To determine attitudes and stereotypes in children, questionnaires in which children are asked to evaluate or assign positive and negative adjectives to figure silhouettes representing different body sizes have been used (Puhl & Latner, 2007). Children have also been asked to report their playmate preference, or to make peer and friendship nominations, and qualitative interviews have also been undertaken. In younger children, attitudes towards
obesity have been assessed by story-telling methods (Cramer & Steinwert, 1998). But, to the best of my knowledge, only one study has assessed implicit attitudes about obesity in children (Solbes & Enesco, 2010). Survey methods have the advantage of elucidating the stereotypes held by individuals concerning obesity, but they are often subjected to social desirability bias and social norms, and do not capture information regarding actual behaviour. Experimental studies, which have the advantage of enabling inferences about causality, have also been performed; for instance, children have been provided with information suggesting that the targets evaluated had little responsibility for their obesity (Bell & Morgan, 2000; DeJong, 1993;
Sigelman, 1991). There are caveats, however, in particular that responses may not necessarily correspond to actual behaviours. Studies of self-reported experiences of teasing and bullying among children with obesity are also scarce, and the studies that have examined obesity stigma as a possible mediator of physical or psychosocial outcomes in children with obesity are still few in number.
Studies of adults have, to a high degree, relied on questionnaires that make certain statements about obese people’s character and behaviour, etc. (Crandall, 1994; Latner, O'Brien, Durso, Brinkman & Macdonald, 2008). But there are also studies that have used adjective ratings (Teachman, Gapinski, Brownell, Rawlins & Jeyaram, 2003) and qualitative designs (Brown, 2006). In the last decade, further studies have been directed at unconscious attitudes or attitudes that people are not aware of or try to deny (Teachman & Brownell, 2001). There are only a few studies that measure stigma by directly asking people with obesity about their experience of discrimination (Puhl & Heuer, 2009), but there have been quite many studies of people’s behaviour or rejection of people with obesity. These, mostly indirect measures, have involved investigating, for instance, seating distance from an obese person, time before a sales person responds to an obese customer, weight penalties, and job applicant ratings. However, field studies that can capture behaviours disclosing when perpetrators and targets interact in the real social world are few (King, Shapiro, Hebl, Singletary & Turner, 2006).
2.5 PREVALENCE AND AETIOLOGY OF OBESITY
The prevalence of obesity has increased in Sweden among both men and women (Lissner, Johansson, Qvist, Rössner & Wolk, 2000; Rasmussen, Johansson & Hansen, 1999). Studies of young men show a five-fold increase in the prevalence of moderate obesity (BMI of 30-34.9) and a ten-fold increase for severe obesity (BMI 35 or over) over a thirty-year period (Neovius et al., 2008). The prevalence of obesity in the Swedish adult population is today approximately 10% in both men and women, according to estimates based on self-reported BMI (Neovius, Janson & Rössner, 2006), and about 25% of the obese group seem to be individuals with severe obesity (Neovius et al., 2008). Recent data from Statistics Sweden show that prevalence between 2000/2001 and 2006/2007 has been stable, at around 10% (Statistics Sweden, 2010).
Results from the US show a similar stabilizing trend in obesity prevalence for a similar period (between 1999/2000 and 2007/2008) in women, while a slight increase in prevalence was recorded for men. The prevalence of obesity in the US is estimated to be 32% for men and 35% for women (Flegal, Carroll, Ogden & Curtin, 2010).
Reports on secular trends in children are scarce from Sweden; however, they seem to correspond to the trend in adults. The prevalence of obesity in children 10 years of age in Gothenburg increased four to five-fold, to 2.9%, between 1983/85 and 2000/2001 (Neovius et al., 2006). A recent study reports that the prevalence of obesity (approximately 3-4%) among 10-11 year-old boys and girls stabilized between 1999 and 2005 (Lissner, Sohlström, Sundblom & Sjöberg, 2009). Such levelling-off in obesity prevalence corresponds to the findings of studies in other countries (Brown, Byatt, Marsh & McPherson, 2009). Sweden ranks low on the international obesity list (World Health Organization, 2010), but – because of a high rate of tracking over time – a majority of children with obesity will remain obese in adulthood (Guo, Wu, Chumlea & Roche, 2002).
Empirical evidence suggests that body weight is determined by complex interaction between biological and environmental factors (Bouchard, 2008), and family studies show that genetics play an important role in obesity (Rasmussen, Magnusson & Sörensen, 2008). For instance, responses to calorific reduction (Hainer, Stunkard, Kunesova, Parizkova, Stich &
Allison, 2000), overfeeding (Bouchard, Tremblay, Despres, Nadeau, Lupien, Theriault et al., 1990), and exercise (Bouchard, Tremblay, Despres, Theriault, Nadeau, Lupien et al., 1994) are more similar within twin pairs than between twin pairs. Heritability is 50-80%, depending on population, age and period of time investigated (Rasmussen et al., 2008). However, it is suggested that the genetic influence has four levels: genetic obesity, strong genetic predisposition for obesity, slight genetic disposition for obesity, and genetically resistant to obesity (Loos & Bouchard, 2003). Slight genetic predisposition to obesity seems to be common in the population, and lifestyle, social, cultural and community factors will therefore be important for its development. Obesity is strongly associated with measures of socioeconomic position, place of residence, ethnicity, age and sex (Kark & Rasmussen, 2005; Neovius &
Rasmussen, 2008b; Ogden, Carroll, Curtin, McDowell, Tabak & Flegal, 2006). Social position in men and attained education in women show a strong association with obesity in most developed countries (McLaren, 2007), even though the disparity in obesity across SES categories appears to have become less (Zhang & Wang, 2004). Hypotheses regarding social differences in obesity levels are several, but SES is likely to be both a causal factor for obesity and a consequence of obesity (Karnehed et al., 2008).
2.6 MEASURING OBESITY
Overweight and obesity are defined as an accumulation of excessive body fat that presents a risk to health. The World Health Organization (WHO) defines overweight, for adults, as having a BMI of 25 to 29.9, and obesity as having a BMI of 30 or over. BMI is calculated as weight in kilograms divided by squared height in metres (kg/m2). The definition is based on the association between BMI and mortality, where studies indicate that the risk of mortality increases from BMI 25 and increases further over BMI 30 (Neovius, Sundström & Rasmussen, 2009; WHO, 2000). BMI has shown to provide a relatively good estimate of body fatness in epidemiological studies (WHO, 1997). However, BMI does not distinguish fat mass from muscle mass, and there is therefore a risk of misclassifying individuals. This is especially salient for men who are more likely than women to have a high BMI because of muscle mass rather
than fat mass. At a group level, however, BMI correlates fairly well with percentage of body fat (Gallagher, Heymsfield, Heo, Jebb, Murgatroyd & Sakamoto, 2000). However, depending on the outcome of interest, different measures of obesity or fatness may be needed (Heitmann, Frederiksen & Lissner, 2004; Rexrode, Carey, Hennekens, Walters, Colditz, Stampfer et al., 1998). In large epidemiological studies, BMI and waist- or hip circumference are quick, easy and cheap measurements of overweight and obesity, but other methods – like underwater weighing, air displacement plethysmography, labelled water techniques and dual-energy X-ray absorptiometry – are the most reliable in obtaining accurate measures of body fat (Parker, Reilly, Slater, Wells & Pitsiladis, 2003; Wang, Deurenberg, Guo, Pietrobelli, Wang, Pierson, Jr. et al., 1998).
For children, the International Obesity Task Force has developed sex-specific cut-offs for overweight and obesity (Cole, Bellizzi, Flegal & Dietz, 2000). BMI in children correlates quite well with body fat, as seen in adults. However, although the cut-offs have been seen to have high specificity (not misclassifying normal weight children as overweight or obese), they are regarded as having low sensitivity (misclassifying overweight and obese children as normal weight) (Neovius & Rasmussen, 2008a). Another way of defining overweight and obesity in children has been to use certain BMI percentiles of specific reference populations (Kuczmarski
& Flegal, 2000).
It is difficult to ascertain at which BMI level the risk of stigmatization increases, because different approaches have been used in the measurement of obesity stigma. Often, a range of body figures representing extreme thinness to extreme obesity has been used, but there have been no known specific BMI levels attached to the figures (Kraig & Keel, 2001; Rand
& Wright, 2000). Recently, there has been some development in these pictorial methods, which include BMI-based body size figures (Harris, Bradlyn, Coffman, Gunel & Cottrell, 2008; Swami, Chan, Wong, Furnham & Tovée, 2008). Furthermore, surveys have asked for people’s attitudes to fat or obese people, or people who weigh too much, without giving participants guidance on the medical definitions, which allow them to make their own characterizations (Allison, Basile
& Yuker, 1991; Crandall, 1994). Conceptualization of an overweight, obese or fat individual might therefore not be consistent across individuals and groups of people. Crandall, for instance, has chosen to use the wording fatness instead of obesity to distinguish between the medical condition (obesity) and a descriptive feature of someone’s body size (Crandall, 1994).
In large surveys that assess perceived discrimination, self-reported BMI has often been used (Puhl, Andreyeva & Brownell, 2008), and in clinical samples or studies in schools, where the sample sizes are often small, objectively measured BMI has been most commonly employed (Friedman, Reichmann, Costanzo, Zelli, Ashmore & Musante, 2005; Koroni, Garagouni-Areou, Roussi-Vergou, Zafiropoulou & Piperakis, 2009; Vartanian & Shaprow, 2008).
2.7 OBESITY STIGMA – CHILDREN
Existing studies of obesity stigma suggest that it is a common experience for children with obesity to encounter negative attitudes and victimization/teasing. However, due to the variety of assessment methods that have been used it is difficult to ascertain the specific prevalence rates of biased attitudes or stigmatizing encounters.
Gender effects
A systematic review of gender and stigmatization of obesity, including both attitudes towards obesity and perceived victimization/teasing among individuals with obesity, indicates that girls are, to a greater degree, subjected to obesity stigma than boys (Tang-Peronard & Heitmann, 2008). This discrepancy seems, however, to be manifested in relation to the type of stigmatization. The sex difference was also more pronounced in studies with older age groups (11-18 years) than younger (2-5 and 6-10 years). Furthermore, studies finding sex differences were often the ones having the highest quality, suggesting that this result may also be due to methodology. A recent prospective study of 8210 7-8-year-old children finds that 36% of boys and 34% of girls with obesity self-report being stigmatized due to their weight (Griffiths, Wolke, Page & Horwood, 2006). Later figure silhouette studies of elementary-school children also show inconsistency in the judgment of the obese target by gender. The target with obesity was more disliked by girls than by boys in one study (Koroni et al., 2009), while boys were the ones making more negative evaluations of a silhouette with obesity in another (Penny &
Haddock, 2007a). Studies including the effects of the sex of both the stigmatizer and the stigmatized are, however, scarce (Cramer & Steinwert, 1998; Klaczynski et al., 2009; Kraig &
Keel, 2001; Stager & Burke, 1982). Kraig and Keel (2001) found variations in 7-9 year-olds’
judgments according to target weight, target sex and sex of the child making the judgment, which indicate that all these factors need to be considered. A study investigating target weight, target sex and rater sex interactions among 10-16 year-olds revealed that female targets with obesity were denigrated more than male targets with obesity, but that there was no effect on the ratings based on the sex of the rater (Klaczynski et al., 2009). The difference between ratings of female and male targets with obesity also increased with age.
Age effects
As previously mentioned, negative attitudes towards obesity are seen in early childhood (Cramer & Steinwert, 1998; Lehmkuhl, Nabors & Iobst, 2009; Margulies, Floyd & Hojnoski, 2008; Musher-Eizenman, Holub, Miller, Goldstein & Edwards-Leeper, 2004; Turnbull, Heaslip
& McLeod, 2000). The evidence concerning the development of stronger, weaker or persistent negative attitudes with age is, however, inconclusive. Higher levels of stereotyping were assigned to an obese target by 5-year-olds compared with 3- and 4-year-olds (Cramer &
Steinwert, 1998), and an increase in obesity stereotyping has been seen from kindergarten up to approximately 4th to 5th grade in a handful of studies (Brylinsky & Moore, 1994; Klaczynski, 2008; Sigelman, Miller & Whitworth, 1986; Wardle et al., 1995). There are, however, studies showing no effect of age (ages 8 to 12) (Tiggemann & Wilson-Barrett, 1998), and decreasing negative attitudes from about ages 5-6 to 10-11 (Penny & Haddock, 2007a; Solbes & Enesco, 2010). In one of these studies it was found that implicit attitudes to obesity were detected among 6-year-olds, and that these remained stable up to age 11. However, explicit and implicit attitudes were only correlated in younger children. Studies from the elementary-school period through adolescence are scarce, but they have shown an increase in stereotyping – from early ages to adolescence in boys (Lerner & Korn, 1972), and from ages around 10 up to 16 in both
females and males (Klaczynski et al., 2009). But a lack of an age effect (ages between 9 to 16) on stereotypes assigned has also been documented (Stager & Burke, 1982).
All the studies conducted have been cross-sectional, and not all have included the possible effects of gender or socio-demographics, which limits conclusions about possible developmental shifts. To the best of my knowledge, there is only one longitudinal study of the development of obesity stereotypes (Davison, Schmalz, Young & Birch, 2008). This showed that 9-year-old girls had decreased their reporting of obesity stereotypes in general by the age of 11, but stereotypes, such as obese individuals are lazy and that it is bad to be obese, did not change. Longitudinal trends in weight-related teasing also suggest that this experience decreases among overweight individuals (overweight and obesity analyzed together) as they make the transition from early to mid-adolescence, 42% to 31% in females and 45% to 20% in males (Haines, Neumark-Sztainer, Hannan, van den & Eisenberg, 2008).
Body weight effects
Own body weight has consistently been shown not to have an effect on attitudes towards obesity in elementary-school children, regardless of method used (Counts, Jones, Frame, Jarvie
& Strauss, 1986; Davison & Birch, 2004; Koroni et al., 2009; Kraig & Keel, 2001; Latner, Simmonds, Rosewall & Stunkard, 2007; Tiggemann & Anesbury, 2000; Wardle et al., 1995). It has been suggested that this evidence means that children with obesity are not protected by their “own group”; that is, in-group favouritism is common when it comes to other stigmatized groups, but this is not seen in obesity. A study of children in pre-school shows similar results, but in some judgmental tasks even stronger stereotyping about obesity was expressed by overweight children than by non-overweight children (Cramer & Steinwert, 1998). In another study, pre-school children’s BMI did not predict assignment of negative adjectives to an overweight figure, whereas the number of positive adjectives attributed decreased with increasing BMI (Rich, Essery, Sanborn, DiMarco, Morales & LeClere, 2008). Additional research among pre-school children shows no effect of own body size on stereotyping (Holub, 2008).
Body image effects
Body image is a multidimensional construct and is suggested to incorporate perceptual, affective, cognitive, evaluative and behavioural components (Smolak, 2004). Pre-school children who perceived themselves heavier showed less obesity stereotyping (Holub, 2008), while, in another study, the numbers of positive and negative adjectives assigned to an overweight figure were not related to perceived body size (Rich et al., 2008). The former authors suggest that this could be a sign of in-group favouritism. Children recognizing themselves as overweight or obese, despite the fact that they are thinner might identify with these children and therefore show less stereotyping. How satisfied one is with one’s body size (domain-specific self-esteem) has also been found to be related to the number of negative stereotypes assigned to a figure with obesity. In a study of 7-12 year-olds, boys’ greater body dissatisfaction predicted more negative stereotypes of a boy figure with obesity. No relationship was seen between girls’ body dissatisfaction and obesity stereotypes assigned to a girl figure (Tiggemann & Wilson-Barrett, 1998). However, among 9-year-old girls, more
interactions with mothers about body shape issues and weight loss predicted a higher level of negative obesity stereotyping (Davison & Birch, 2004). However, studies that have taken into account both target sex and rater sex in relation to body image and body stigmatization are scarce.
Media effects
A systematic review shows that the media present obese people in a biased manner (Greenberg et al., 2003) and, as mentioned previously, the stigmatization of obese individuals in a Swedish media context has been demonstrated (Sandberg, 2007). Advertisements and news media are also common arenas for framing messages that emphasize personal responsibility for weight and obesity (Puhl & Heuer, 2009). In conjunction with the denigration of obese individuals, magazines and television strongly communicate the thin ideal (Thompson & Heinberg, 1999), and in children’s literature and movies, characters with obesity are often negatively presented (Puhl & Heuer, 2009). Therefore, it is likely that the media transmit negative stereotypes about obesity to children. Despite this, only two published studies have investigated the association between media use in children and obesity stigmatization. Among elementary-school children, boys’ higher television viewing predicted increased stereotyping of obese girls (Harrison, 2000).
However, boys’ and girls’ interpersonal attraction to a male television character with obesity predicted less stereotyping of a girl and boy figure with obesity. Elementary-school children’s weekly playing of video games and total media use have also been associated with obesity stigmatization (Latner, Rosewall & Simmonds, 2007). In boys, total television viewing was correlated with obesity stigma, while girls’ magazine reading was an independent predictor of obesity stigma. Studies of undergraduate females suggest that time spent reading fashion magazines may be related to negative attitudes towards obesity through dysfunctional appearance beliefs (Lin & Reid, 2009).
Parental effects
There is little research on parental influence on children’s likelihood of expressing obesity stigma. However, previous research has found that fathers’ and mothers’ higher investment in physical appearance predicts higher obesity stereotyping. Fathers and mothers emphasizing thinness and weight loss in interactions with their daughters also predict higher levels of obesity stereotyping in their 9-year-old girls (Davison & Birch, 2004). No direct link was found in the study between parents’ obesity stereotypes and their daughters’ obesity stereotypes.
However, a study of adolescents has documented agreement between girls’ obesity prejudice and their mothers’ obesity prejudice (O'Bryan, Fishbein & Ritchey, 2004). Mothers’ higher BMI and higher body dissatisfaction were independent predictors of pre-school boys’ and girls’
higher numbers of negative adjectives assigned to a figure with overweight (Rich et al., 2008).
Furthermore, parents asked to tell a story to their children of a child of average weight, one with a disability and one with obesity, imaged the child with obesity as having the lowest self- esteem and being least successful (Adams et al., 1988). Studies also show that adolescents with obesity may be stigmatized by their own parents, e.g. parents having paid for college for their non-obese child, but not for their obese child (Crandall, 1995). There seems to be only one study
that has investigated parent’s controllability beliefs regarding weight. This demonstrated that parents of 4-6 year-olds who held higher controllability beliefs on weight were more likely to restrict their children’s food intake (Musher-Eizenman, Holub, Hauser & Young, 2007).
Effect of the belief in personal responsibility for weight
It has been suggested that the belief that weight is controllable, or that people are personally responsible for their own weight, is one of the strongest reasons for people’s negative responses to obesity (Crandall, 1994). Such a belief is also correlated with negative attitudes towards or stereotypes about obesity in pre-school children (Lehmkuhl et al., 2009; Musher-Eizenman et al., 2004) and in elementary-school children (Tiggemann & Anesbury, 2000). But these studies are correlational, and studies using an experimental design that have tried to combat negative attitudes towards obesity by using this notion have shown only modest effects. Children aged 6 and 10 who were given information that the obese target had little personal responsibility were less likely to blame the target for the condition compared with children who were not given any information about cause (Sigelman, 1991), but the children did not show more positive responses afterwards. Similar findings have been documented in other studies among 9-11 year-olds; less blame was assigned to the obese target if a medical (Bell & Morgan, 2000) or a genetic (Anesbury & Tiggemann, 2000) explanation was given. In the study by Bell and Morgan (2000), younger children were found to decline in their negativity, while older children did not change. This was also seen in the study by Anesbury and Tiggemann (2000). Adolescent males did not reduce their obesity stigmatization when a medical explanation for obesity was given, while females were less likely to display stigma when given such information (Kingsbury, 2009). One recent study has shown that internal causal attributions for obese targets increases with age (from 10 to 16) and that attributions become much stronger with age for female targets than for male targets (Klaczynski et al., 2009).
2.8 OBESITY STIGMA – ADULTS
Adults’ negative attitudes towards obese individuals have been seen to be as pervasive as those of children (Puhl & Brownell, 2001). Increasing prevalence of weight discrimination has also been documented over the last decade (Andreyeva, Puhl & Brownell, 2008) in parallel with the trend of increasing obesity. When it comes to discrimination due to obesity, both institutional (health care, employment, education) and interpersonal mistreatment have been reported (Puhl
& Heuer, 2009), and obesity stigma may be higher for women (Andreyeva et al., 2008).
Personal responsibility for weight
Studies among adults that give information about the uncontrollability of obesity seem to be more effective in reducing stigma than similar studies among children (Crandall, 1994; DeJong, 1993; Grosko, 2007; Teachman et al., 2003; Weiner et al., 1988). However, the interventions are often very short and highlight different aspects of controllability (Danielsdottir et al., 2010). In their interactions, people do not normally have information regarding individual causes of obesity, so when access to this information is lacking, the person with obesity is assumed to be directly responsible for his or her condition (Ross, Shivy & Mazzeo, 2009). However,
individuals with obesity who are engaged in risk-reducing behaviour, such as exercise, are judged more positively, which suggests that people consider both aetiology and the risk the person with obesity is taking. Also, people with obesity engaged in binge eating are more negatively judged than persons without any indication of a possible blameworthy behaviour (Bannon, Hunter-Reel, Wilson & Karlin, 2009). If there is justification for stigmatization, then it seems more likely to be demonstrated (Crandall & Eshleman, 2003), however, if there is ambiguous responsibility for obese people’s condition the degree of denigration may be similar as that found for people that are clearly responsible for their condition (Ross et al., 2009).
There is ample evidence that men show greater stigmatization of obesity than women (Crandall, 1994; Hebl, Ruggs, Singletary & Beal, 2008; Latner et al., 2008; Lewis, Cash, Jacobi &
Bubb-Lewis, 1997; O'Brien, Latner, Halberstadt, Hunter, Anderson & Caputi, 2008; Puhl et al., 2005; Wang, Brownell & Wadden, 2004), but there is no evidence that men show higher belief in the controllability of weight (Bannon et al., 2009; Crandall, 1994; Lewis et al., 1997; Puhl et al., 2005; Teachman et al., 2003). Furthermore, adults’ own body weight has not been found to correlate with controllability beliefs (Crandall, 1994; Lewis et al., 1997; Puhl et al., 2005; Quinn
& Crocker, 1999; Wang et al., 2004). However, higher BMI among women predicts a weaker implicit attitude that people with obesity are lazy (Teachman et al., 2003), which – in parallel with research using explicit measures – shows that individuals with obesity are less likely than thinner individuals to regard people with obesity as lazy (Ross et al., 2009; Schwartz, Vartanian, Nosek & Brownell, 2006). Numerous studies though show that one’s own body weight is seldom related to explicit attitudes or stereotypes about people with obesity (Puhl & Brownell, 2003).
Work-life
Employees with overweight and obesity have been viewed as less conscientious, less agreeable, less emotionally stable and less extraverted than normal weight employees (Puhl & Heuer, 2009). A study using a nationally representative sample of adults, however, challenges this view, and shows that personality and body weight are very weakly associated, and that age and sex are more important (Roehling, Roehling & Odland, 2008). Two recent meta-analyses of experimental studies of obesity stigma in the workplace, however, conclude that overweight individuals are likely to be disadvantaged across different workplace outcomes compared with their normal weight counterparts (Roehling, Pilcher, Oswald & Bruce, 2008; Rudolph, Wells, Weller & Baltes, 2009). The effects of obesity stigma on hiring were not different between managerial and sales persons in one of the studies (Rudolph et al., 2009), but in the other study overweight employees were found to be more disadvantaged when it came to be evaluated for jobs with greater public contact (Roehling et al., 2008). In the meta-analysis by Roehling and colleagues (2008), women and men were equally likely to be subjected to discrimination. A majority of the studies reviewed were laboratory studies and not field studies. Studies of perceived employment discrimination are few, but women seem to have a greater tendency to report weight-related discrimination than men, which is apparent in all weight groups from normal weight to severe obesity (Puhl et al., 2008; Roehling, Roehling & Pichler, 2007); by
